1. Very little is known about the biological availability of manganese from human milk and other infant milk diets. To determine the relative Mn availability, and to examine whether the age and the duration of previous fasting affect Mn absorption, sucking rats were given human milk, bovine milk and infant formula (regular Similac; Ross Laboratories, Columbus, OH) extrinsically labelled with 54Mn.2. Milk diets were given by gastric intubation and the radioactivity of the carcass, liver and digestive tract was measured 3 h after feeding.3. The concentration of endogenous Mn was lowest in human milk (7-10,ug/l) and highest in rat milk (1W165 ,ug/l). Increasing the non-radioactive total Mn concentrations of either human milk or bovine milk up to 150 pg/l did not affect the absorption of 54Mn by 10-d-old rats.4. No significant ( P > 0.05) difference in 54Mn absorption was found among the three milk diets (human milk, bovine milk, infant formula) in 8-to 1 1-d-old rats. However, significantly more ( P < 0.05) 54Mn was absorbed from human milk and infant formula than from bovine milk when 13-d-old rats were used.5. 54Mn radioactivity detected in carcasses of 8-, 9-, 10-and 1 1-d-old rats ranged from 25 to 27% of the dose frbm various milk diets. The activities of 54Mn in the carcasses of 13-d-old rats were 15, 11, and 16% of the dose from human milk, bovine milk and infant formula respectively.6. The trend of 54Mn incorporation into liver was similar to that of the carcass and over 60% of the absorbed 54Mn was incorporated into the liver regardless of the type of milk used.7. Absorption of 54Mn from extrinsically labelled rat milk using 9-or 10-d-old sucking rats was similar to its absorption from infant formula.8. The absorption of 54Mn from the three milk diets decreased with age of the neonatal rats and 54Mn absorption from human milk, bovine milk, infant formula as well as rat milk was affected similarly by duration of previous fasting.